We have updated our ESP Muse library to support OpenAi Chatgpt Voice application
You have the freedom to choose between the two architectures:
Demo 1: Realtime API using WebSockets
Ultra-fast response times, ideal for interactive voice assistants.
Demo 2: Chained architecture — processes audio step by step:
Audio → Text → LLM response → Text-to-Speech
A little slower (2–3 seconds of latency), but extremely cost-effective to run.
Unlike many examples you can find online that require a local or cloud server to do all the heavy lifting, this library allows the ESP32 itself to communicate directly with the OpenAI API. No intermediate server, no complicated setup — just your device and the API.
The big advantage of this library is that it works directly with Arduino — no need to use the ESP-IDF, which can often be complex and time-consuming to set up. With this library, you can prototype and run AI-powered voice applications on your ESP32 in just a few minutes.
This library is compatible with the full ESP Muse product line:
- Muse Luxe
- Muse Radio
- Muse Proto
- Muse Manga
Our hope is that this release will help you experiment with new AI-driven audio applications — projects that could truly transform the way we interact with technology in our daily lives.